Skip to main content

Full text of "Effects of aerosols on the determination of the temperature of the earth's surface from radiance measurements at 11.2 [micro]m / H. Jacobowitz, K.L. Coulson"

See other formats



NOAA Technical Report NESS 66 


National Oceanic and Atmospheric Administration 

National Environmental Satellite Service 

Effects of Aerosols on the Determination 
of the Temperature of the Earth's Surface 
From Radiance Measurements at 11.2 ^m 


September 1973 


National Environmental Satellite Service Series 

The >nmental Satellite Service (NESS) is responsible for the establishment and operation 

ational Meteorological Satellite System and of the environmental satellite systems 

The three principal offices of NESS are Operations, Systems Engineering, and Research. The 

Technical Report NESS series is used by these offices to facilitate early distribution of research 

results, data handling procedures, systems analyses, and other information of interest to NOAA organiza- 

Publication of a report in NOAA Technical Report NESS series will not preclude later publication in an 
-ided or modified form in scientific journals. NESS series of NOAA Technical Reports is a continua- 
tion of, and retains the consecutive numbering sequence of, the former series, ESSA Technical Report 
:nal Environmental Satellite Center (NESC), and of the earlier series, Weather Bureau Meteorological 
: Llite Laboratory (MSL) Report. Reports 1 through 57 are listed in publication NESC 56 of this ser- 
ies . 

Reports 1 through 50 in the series are available from the National Technical Information Service 

(NTIS), U.S. Department of Commerce, Sills Bldg. , 5285 Port Royal Road, Springfield, Va. 22151. Price 

3 paper copy; $1.45 microfiche. Order by accession number, when given, in parentheses. Beginning 

51, printed copies of the reports are available through the Superintendent of Documents, U.S. 

Government Printing Office, Washington, D.C. 20402. Price as indicated. Microfiche available from NTIS 

(use accession number when available). Price $1.45. 

ESSA Technical Reports 

58 Angular Distribution of Solar Radiation Reflected From Clouds as Determined From TIROS IV Radi- 
ometer Measurements. I. Ruff, R. Koffler, S. Fritz, J. S. Winston, and P. K. Rao, March 1967. 

NESC 59 Motions in the Upper Troposphere as Revealed by Satellite Observed Cirrus Formations. H. 
McClure Johnson, October 1966. (PB-173-996) 

NESC 40 Cloud Measurements Using Aircraft Time-Lapse Photography. Linwood F. Whitney, Jr., and E. Paul 
McClain, April 1967. (PB-174-728) 

NESC 41 The SINAP Problem: Present Status and Future Prospects; Proceedings of a Conference Held at 
the National Environmental Satellite Center, Suitland, Maryland, January 18-20, 1967. E. Paul 
McClain, October 1967. (PB-176-570) 

42 Operational Processing of Low Resolution Infrared (LRIR) Data From ESSA Satellites. Louis 
Rubin, February 1968. (PB-178-123) 

4 5 Atlas of World Maps of Long-Wave Radiation and Albedo--for Seasons and Months Based on Measure- 
ments From TIROS IV and TIROS VII. J. S. Winston and V. Ray Taylor, September 1967. (PB-176- 

44 Processing and Display Experiments Using Digitized ATS-1 Spin Scan Camera Data. M. B. Whitney, 
R. C. Doolittle, and B. Goddard, April 1968. (PB-178-424) 

5 The Nature of Intermediate-Scale Cloud Spirals. Linwood F. Whitney ,■ Jr . , and Leroy D. Herman, 
May 1968. (AD-673-681) 

NESC ithly and Seasonal Mean Global Charts of Brightness From ESSA 3 and ESSA 5 Digitized Pic- 
tures, February 1967-February 1968. V. Ray Taylor and Jay S. Winston, November 1968. (PB-180- 

nomial Representation of Carbon Dioxide and Water Vapor Transmission. William L. Smith, 
1969. (PB-183-296) 

.tistical Estimation of the Atmosphere's Geopotential Height Distribution From Satellite 
nation Measurements. William L. Smith, February 1969. (PB-183-297) 

optic/Dynamic Diagnosis of a Developing Low-Level Cyclone and Its Satellite-Viewed Cloud 
Patterns. Harold J. Brodrick and E. Paul McClain, May 1969. (PB-184-612) 

timating Maximum Wind Speed of Tropical Storms From High Resolution Infrared Data. L. F. 
Ilub€ Timchalk, and S. Fritz, May 1969. (PB-184-611) 

(Continued on inside back cover) 



Frederick B. Dent, Secretary 

Robert M. White, Administrator 

David S. Johnson, Director 


NOAA Technical Report NESS 66 

Effects of Aerosols on the Determination 
of the Temperature of the Earth's Surface 
From Radiance Measurements at 11.2 ^m 

H. Jacobowitz 
K. L. Coulson 




September 1973 

UDC 551.521.2:551.510.42 

551.5 Meteorology 

.510 Atmospheric structure and composition 

.42 Impurities and dust 

.521 Radiation 

.2 Terrestrial radiation 

For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C., 20402. 

Price $0.55 



Abstract I 

1. Introduction I 

2. Atmospheric models 2 

3. Radiative computations 4 

4. Results of the calculations 8 

A. Extinction and scattering by aerosols 8 

B. Aerosol effects on temperature determinations 10 

Acknowledgements 17 

References 18 


Digitized by the Internet Archive 

in 2013 


H. Jacobowitz 

National Environmental Satellite Service 
Washington, D. C. 


K. L. Cou I son 

University of California 
Davis, Ca I i fornia 

ABSTRACT. The influence of atmospheric aerosols on out- 
going terrestrial radiation in the window region at 
11.2 ym is estimated by atmospheric modeling. Nine 
different particle size distributions, 3 vertical aerosol 
concentration profiles, and 5 vertical temperature profiles 
are considered, for a total of 135 atmospheric models. For 
each model, the upward radiant intensity was computed for 
every 0.2 km from the surface to 20 km. From these values, 
equivalent black-body temperatures were determined as a 
function of height, and the error in the derived surface 
temperature due to aerosols was computed. Results show 
that absorption by aerosols is dominant over emission, 
making the deduced surface temperature lower than actual, 
except in the subarctic winter. Errors were generally 
less than 2°K except under extreme conditions. 


The influence that aerosols (atmospheric particulate matter) have on the 
radiation emitted to space from the earth-atmosphere system has come under 
increasing study in the past few years, particularly with regard to the 
climatic changes that may be taking place as a result of the increase in the 
pollution levels in our atmosphere. While such studies are of great value, 
the radiative studies are also important in assessing the effects of aerosols 
on the radiometric measurements made routinely from earth-orbiting satellites. 

For example, vertical temperature profiles are deduced from radiometric 
measurements made in several narrow spectral intervals in one or more atmos- 
pheric absorption bands. However, these deductions require an accurate know- 
ledge of the transmission properties of the atmosphere in those bands. While 
the transmission trhough the principal atmospheric gases is accounted for, 
the transmission through the aerosols is generally neglected. It is therefore 
of interest to study how this neglect affects temperature determinations and 
to learn how one can compensate for aerosol effects. 

It was for these purposes that the present study was initiated. Although 
emphasis will be on possible errors that may be produced in the surface 
temperature by a neglect of the effects of aerosols, the results will be 

applicable to certain other problems as well. 

Recently, Stowe (1971) investigated absorption and scattering by aerosols 
of radiation in the various spectral 'regions currently used for atmospheric 
sounding. It was found that the intensity of the radiation scattered by the 
aerosols is small compared to that emitted. Thus we assume that accuracy 
will still be sufficient if scattering is neglected. Stowe used the "Haze M" 
particle size distribution of Deirmendjian (1969) and the vertical profiles 
of particle concentration obtained by Rosen (1967) and Elterman (1968). He 
estimated that radiance determinations may be in error by as much as \5% in 
the most transparent part (13.33 ym) of the 15 ym band of carbon dioxide 
because of a neglect of absorption by aerosols. He also found that most of 
the radiation absorbed from the upward stream is compensated for by radiation 
emitted from the aerosol particles themselves, thereby reducing errors in 
radiance estimates to generally less than \%. According to Stowe's analysis, 
errors in the remote sensing of the temperature profiles are due to ( I ) large 
underestimates of the atmospheric emission and (2) errors in the weighting 
functions used for Inverting the radiometric measurements caused by a neglect 
of the effects of atmosphertc aerosols. 

In a recent study of aerosol effects in the natural atmosphere Mariatt 
et al. (1971) measured atmospheric aerosol size distribution and sea surface 
temperatures from altitudes ranging from 400 to 37,300 feet. Using the 
measured aerosol concentrations and nearby radiosonde temperatures, he was 
able to determine that for very hazy conditions with cloudless skies, 
error in the computed blackbody temperature in the 8- to 14 ym window region 
due to neglect of the observed aerosol absorption was of the order l°K. 

For the present study we have focused our attention on estimating theoreti- 
cal ly the effect of aerosols on the I I -ym window region. This region was 
chosen to minimize the perturbing effects of absorption by atmospheric gases 
and to estimate what accuracy of surface temperature determinations can be 
reasonably expected from satellite measurements. The analysis can be readily 
generalized to include other spectral regions. 


Calculations were made for a variety of model atmospheres. Since we were 
chiefly interested in computing the difference in upward intensities when 
aerosols were included and when aerosols were excluded, it was assumed that 
the gaseous component of the atmosphere could be omitted from the models with- 
out introducing serious errors. This is equivalent to the assumption that 
the effect of the gases upon the emitted radiation is independent of the 
effect of the aerosols, i.e., their combined effects are additive. Although 
this is probably a poor assumption for many atmospheric absorption bands, it 
should be a good one for the window region because of the weak molecular 
absorption in that region. 

Natural aerosols generally range in radius from 0.001 to 100 ym although 
the greatest concentrations of the particles are generally between 0.1 and 
10 ym. When there are dust storms, fog, or smog, the distribution is 
changed somewhat by the addition of particles generally exceeding I ym in 

size. Junge (1953) found that the observed size distribution could be 
described reasonably well by the expression 

d^ = C r 



where N is the particle concentration, r is the particle radius, and C is a 
proportionality constant. Later investigations by Junqe (1955) showed that 
(I) could be generalized as 

dN = C r" (2) 


where v, generally known as the Junge distribution parameter, is only approx- 
imately equal to 4. Variations of v from 4 occur as a result of humidity 
changes, dust storms, or other variations in atmospheric phenomena. 

To explore the influence of the particle size distributions on the 
radiation transferred by the atmosphere, three values of v (3.5, 4.0, and 
5.0) and three size ranges (0.08 to 10.0 ym, 0.06 to 5.0 ym, and 0.04 to 
3.0 ym) were chosen since their scattering and extinction cross sections were 
available (I. L. Zel'manovich and K. C. Shifrin, 1971). 

To distinguish among clear, hazy, and very hazy atmospheres, vertical 
profiles of aerosol number density based upon those of McClatchey et a I . 
(1971) were used. These are shown in figure I, where the upper limit chosen 
is an altitude of 20 km. The aerosol concentration above this level in their 


— r ~ V 

1 1 1 II II 

1 !~ 

i i i i ii| 1 — i i i i i ii| 1 — i i i i 1 1 1 


























(Sfc. Vis. -5 km| 






CLEAR „-- 
(Sfc. vis. = 





\^^^,~-~~''(Sfc. Vis. = 1 km) 

1 E^ iT^ ■ T-l-t^j 1 1 


10 10 J 10 10 

Particle Concentration (number per cm ) 

Figure I. — Vertical profiles of aerosol particle concentration for clear, 
hazy, and very hazy atmospheric conditions. 

models is too low to have any significant effect upon the radiation field. 
For the clear and hazy atmospheres, the number density at the surface is such 
that the surface visibi I ities for the case that 0.08 ^_ r <_ 10.0 ym and 
v = 4.0 are 3 and 5 km, respectively. A model of a very hazy atmosphere was 
constructed from that of the hazy atmosphere by arbitrarily extrapolating 
the concentration of the hazy atmosphere at the I -km level to a surface value 
corresponding to a visibility of I -km. 

The models used for the aerosol distributions then were specified completely 
by assuming that the relative size distributions were independent of altitude. 
Bu combining the 9 relative size distributions with the 3 vertical concen- 
tration profiles, we obtained 27 models of the aerosol distribution. 

All that remains to be specified in the atmospheric models are the vertical 
temperature profiles. Again, following McClatchey et al, (1971), we chose 5 
profiles corresponding to mean conditions in the tropics and to the summer 
and winter seasons for the mid latitudes and subarctic regions (fig. 2). Each 
of the 5 temperature profiles were combined with each of the 27 aerosol 
profiles to make a total of 135 model atmospheres for which computations were 


The scattering and extinction cross sections of the aerosols tabulated by 
Zel'manovich and Shifrin were computed using the complex index of refraction 
corresponding to that for water at a temperature of 20°C. The extinction 
and scattering properties for the different aerosol models at a wavelength 


Figure 2. — Vertical profiles of the mean temperature for the tropical region 
and for the summer and winter seasons of the mid latitude and 
subarctic regions. 

A • 112/um, m = 1.133 - 0.124i 

Extinction Cross Section 
Scattering Cross Section 
Albedo of Single Scattering 

0.08 ^ r ^ 10.0 

0.06 ± r i 5.0 

0.04 i r ^ 3.0 

0.08 i r ^ 10.0 

. v 0.06 - r i 5.0 

0.04 £ r i 3.0 


40 45 

Junge Distribution Paromet 

Figure 3. — Extinction and scattering cross sections and the albedo of single 
scattering vs. the Junge distribution parameter (v) for the 
wavelength (A) and index of refraction (m) indicated. Size ranges 
of the particles are given to the right of each curve. 

equal to 11.2 ym can be seen from the plot of effective cross section 
(extinction: o ex -\-; scattering; cr ex 4.) vs. the Junge exponent v (fig. 3). The 
results have been normalized by dividing the effective cross section for the 
aerosols by the total' number of particles assumed, which yields an effective 
cross section of one particle which is radiatively typical of the total 
aerosol model . 

One of the basic parameters that characterize the transfer of radiation by 
the aerosols is the extinction optical thickness t^ of a layer. This is 
defined to be the property of a layer that causes unit radiation of wave- 
length X normally incident upon it to be reduced to e~^X in passing through 
the layer. It may be computed from a knowledge of the extinction cross 
section of the aerosols and their vertical concentration profile by means of 
the equation 

t x (h) 





where a, . is the normalized extinction cross section at the wavelength \ 
for a unit aerosoi concentration, N(Z) is the concentration as a function of 
the altitude Z and h is the altitude of the top of the aerosol layer. The 
cross sections given by Zel'manovich and Shifrin and the vertical concentra- 
tion profiles shown in fig. I were used in eq. (3) to compute the optical 
thickness as a function of the altitude for all 27 aerosol models. Table I 

TABLE I. — Optical thickness of model aerosol atmospheres 

r ( m) 

r ( m) 




icai thic 




Very hazy 
















































0. 124 







shows the resulting optical thickness for the models for the layer extending 
from the surface to the 20*-km level. 

From the variation of the optical thickness with altitude, the transmissi- 
vity from any level in the model to any other level could be readily computed 
By restricting the study to a narrow spectral interval and assuming that 
scattering is small compared to absorption (a reasonable assumption for most 
of the aerosol models) one can approximate the transmiss ivity in the vertical 
from the altitude Z| to the altitude Z? by the expression 

(Z „ZJ = e 

-Jt x (Z^ - t x (Z j) sec 6 

X v " I 


where t x (Z ( ,Z 2 ) is the transmiss ivity, t x (Zj) and t x (Z 2 ) are the optical 
thicknesses from the surface to the levels Z| and Z 2 , respectively, and 9 is 
the angle between the direction of the radiation and the vertical. For the 
present study, the radiation was considered only in the vertical direction, 
in which case sec 9 = 1.0. 

The intensity of radiation at any altitude h may be computed from the 

l x (h) 

e A B A ( W° 

,h> + j b a (t(Z)J 

d T (Z,h) 




where I, (h) is the intensity of the upward radiance at the altitude h, 

B^(T) is the blackbody radiance at the temperature T, T is the surface 
temperature, t^(Z,h) is the transmissiyity from the altitude Z to the 
altitude h, e^ is the surface emissivity, and x is wavelength. Since the 
temperature and transmissi vi ty are given at a finite number of levels, (5) 
was replaced by the approximate expression; 

I- 1 

m*0 j H5 L ( a m+ |- i x m' I 


where Z represents the altitude of the £-f- h level and e^ has been set to 
unity for all model atmospheres. 

The equivalent blackbody temperature, T , that corresponds to the radiant 
intensity IMZ ) is defined by the expression, 

Bx'V = ^V ' (7) 

Comparing this with the surface temperature T , the temperature error AT 
is defined by the relation 

AT = T - T 

o eq 


This quantity, which is of primary interest in the study, represents the 
error in the determination of the temperature from a measurement of the 
radiance at some altitude above the surface. In the absence of any aerosol 
or molecular components to the atmosphere, T eq = T Q and therefore AT = 0. 

With aerosols and air molecules in the atmosphere, T ^ T , due to the 

r eq o' 

absorption and emission that takes place in the atmosphere. 

The transport of radiation through the model atmospheres was computed and 
analyzed by means of eqs. (4) to (8). The procedure used was: 

(1) Compute the optical thickness as a function of altitude for every 
0.2 km from the surface to an altitude of 20 km for each of the 27 
aerosol models using (3). 

(2) Determine the transmissivi ties between each of the levels for each 
aerosol model using (4). 

(3) Compute the intensity of radiation directed vertically upward at 
each level for each aerosol model and for each of the 5 temperature 
profiles (135 model atmospheres) using (6). 

(4) Calculate the equivalent blackbody temperature T QQ from the upward 
radiation intensity values by means of (7). 

(5) Determine the temperature errors by (8), using T as determined 
from the radiance values and T as assumed in the atmospheric 
model s. 


A. Extinction and Scattering by Aerosols 

The general extinction and scattering characteristics of the different 
aerosol models can be seen in the curves of effective cross section (fig. 3). 
The effective cross section corresponds to that for a single effective 
particle representative of each distribution. 

The curves for the extinction cross section a ex f show there is a relatively 
strong dependence of extinction cross section on the range of the radii taken 
into account. For instance, cr ex f for 0.08 _< r <_ 10.0 um is about an order of 
magnitude greater than that for 0.04 <^ r <_ 3.0 um, and the ratio between 
these extinction cross sections is practically independent of v. Physically, 
an increase of the minimum size from 0.04 to 0.8 ym effectively eliminates 
many small particles from consideration, since the Junge distribution contains 
many small particles. Conversely, an increase of the upper size limit does 
not add many particles, but those added are very effective attenuators. The 
net result, as the size limits for the radii increase, is an increase in the 
size of the average particle of the distribution and an increase in the value 
of the extinction cross section. 

The same reasons ing can be applied to the curves of the scattering cross 
section, a sea -f.. However, for a given value of v, the rate of increase of 
°scat w '+ n increase in the limits of the radii is greater than that for o ex f. 
This can be explained with the aid of figure 4 in which the extinction and 
scattering efficiency factors (Q ex f =a ext^ 7rr ^ anc ' ^scat =a ext / 7Tr ^) are 
plotted versus the particle radius r for three different values of the complex 
index of refraction (Q ex + > Qscat^ • While the real part of the index of re- 
fraction is the same for all three cases, the imaginary part differs. The 
index for the middle curve of each set is the same as that used in obtaining 
the results of figure 3. By comparing the curves in figure 4 with the super- 
imposed straight lines that have geometric slopes equal to I and 4, it can be 
concluded that a sca ^/i\r^ * r^ and Cg^/frr^ ^ r for all but the largest values 
of the particle radius. Then, cr sca -j- ^ r° and tf ex -|- ^ r^ for sufficiently small 
values of r. The scattering cross section, therefore, is particularly 
sensitive to the sizes of the scattering particles, and anything that changes 
the average size of the distribution has more effect on c sca -f- than on cr ex -j-. 
This property is responsible for the large spread of the curves of figure 3 
for the three size ranges. 

The dependence of the scattering and extinction cross sections on the radius 
also explains the decrease in the value of the cross sections with increase in 
the Junge exponent v. By increasing the magnitude of v, a steeper negative 
slope is given to the distribution. This increases the number of small 
particles at the expense of the large ones and therefore decreases the size of 
the average particle causing the cross sections to decrease. Again, because 
of the greater sensitivity of a , to particle size than <j -., the curves for 



— i — i — i i i 1 1 1 1 1 — i i i i 1 1 1 1 — i — n 

Index of Refraction 
1.15 -0.20i 

1.15 -0.1143i 

1.15 -0.05i 

>// / ft 

0.1 1.0 

Particle Radius (^.m) 

j i i '''I' 


Figure 4. — Extinction and scattering efficiency factors Q(=a/irr ) versus the 
radius r for the wavelength A and index of refraction (m) 
Solid lines that have geometric slopes of I and 4 are superimposed 
for the purpose of comparison. 

a scat decrease more rapidly with increasing v than the curves for (Text* For 
instance, in figure 3 the curve for the middle size range shows a decrease 
in a S cat by a factor of 180 from v = 3.5 to v = 5.0, but a decrease of only 
12.5 for a ex+i 

Since the total attenuation is produced by a combination of absorption and 
scattering, we can write for the respective cross sections the relation 

a ext a abs a scat 
By dividing by cr ex ^- we obtain the quantity 


CT scat q abs 




which is designated the albedo of single scattering, w. Physically, it 
represents the part of the total attenuation that is contributed by scatter- 
ing. Then the quantity 

■ - _ c?abs 
a ext 


represents the part due to absorption. Curves of u" for the various aerosol 
models are plotted, according to the scale given on the right, as the bottom 
family in figure 3. It is striking that such a small fraction of the total 
attenuation is due to scattering. The curves show that the contribution of 
scattering to the extinction is generally between I % and 10 % of the total. 
Only for particles in the size range 0.08 _f_ r <_ 10.0 ym and v = 4.0 is 
w > 10/6, and It gets as low as 0.2 % in the size range 0.04 ± r <_ 3.0 ym. 

There are two consequences of the existence of such smal I values of U for 
the present problem. First, absorptivity of the aerosol particles is high, 
which means if Kirchhoff;s law is assumed valued for aerosols, that their 
emissivity is likewise high. Thus, most of the energy subtracted from a beam 
of radiation traversing the aerosol medium is absorbed by the aerosol 
particles, but their high emissivity ensures that they also lose energy 
efficiently by remitting the radiation. In fact, if the particles were at 
the same temperature as the source from which they were receiving radiation 
(the surface in this problem), they would maintain radiative equilibrium and 
would undergo no temperature change. This does not necessarily mean, however, 
that the same amount of radiation would be returned to the original beam, as 
the angular dependence of the absorbed and emitted radiation may not be the 

The second result of the high absorptivity of the aerosol particles is that 
the single scattering assumption is a good approximation of reality. Because 
of single scattering the radiative field is weak; the series attributable to 

secondary and higher order scattering of already scattered radiation 
converges rapidly. If 10 % is due to primary scattering, only I % would be 
due to secondary scattering,- 0.10 % to tertiary scattering, etc. Thus, any 
errors caused by neglect of multiple scattering would be minor compared to 
those of the unknown physical validity of the models. 

The relatively strong decrease of to with increasing v and the close spacing 
of the three curves of nr shown in figure 3 both indicate the high absorptivity 
of small particles in comparison with that of large particles. As pointed 
out above, cr sca -f- ^ r& and a ex f ^ r- 5 for smal I particles. Since cr ex f is 
dominated by o a ^ s , one can also write o a ^ s *v r^. Thus, as particle size 
decreases, Q sca f decreases about I0- 5 times as fast as a bs> thus producing 
the result shown. 

B. Aerosol Effects on Temperature Determinations 

Using the atmospheric models discussed above, computations have been made 
for determining the error that would be produced by various aerosol distri- 
butions in the intervening atmosphere in surface temperature determinations 
from satellite altitude. The temperature error AT is defined here as the 
difference between the actual temperature, i.e., that assumed to exist at the 
surface, and the equivalent blackbody temperature which would be deduced from 
measurements of the upward radiance at the top of the atmosphere. Thus a 
positive value of AT indicates the actual temperature exceeds that measured, 
and vice versa. 

The temperature errors determined by this procedure are plotted as a function 

of the Junge distribution parameter v for the five models of atmospheric 
temperature profile in figures 5 to 9. In each case, computations of AT have 
been made for three different intervals of particle size and for clear, hazy, 
and very hazy atmospheric conditions. 

Results for the tropical atmosphere are shown in figure 5. The curves are 
defined by three points each, the points corresponding to v 3.5, 4.0, and 
5.0. As would be expected from the results shown previosuly in figure 3, the 
largest temperature errors are for the largest aerosol size range 
(0.08 <_ r _< 10.0 urn), for v = 3.0, and for the very hazy atmosphere. For this 
worst case AT approaches 2.0°K. 

Just how realistic this worst case is for actual atmospheric conditions is 
hard to say. We do not know enough about the aerosols that exist in the 
atmosphere, particularly in the tropical atmosphere, to make a firm judgment. 
However, the very hazy atmospheric model, with its low horizontal visibility, 
does not occur freq uently in the tropics. The clear model is probably more 
applicable for tropical atmospheres, and AT for the worst clear case is only 
about l°K. If the Junge distribution parameter v has a more moderate value 
of 4.0, as has been found to be the case in many instances, then AT for the 
clear tropical atmosphere is only about 0.25°K. 

Throughout the computations, it is shown that AT is very sensitive to the 
range of particle sizes assumed. The magnitude of AT decreases by a factor 
of about 10 as the size range assumed decreases from 0.08 < r <_ 10.0 ym to 
0.04 <_ r <_ 3.0 ym. This means that the error comes within the noise of 
measurements for the smaller size range, but that the result of aerosols of 






\i 0.08 ^ r 

£ 10.0/um 


* V> 

\ V 

\ \\ 

\ V 

\ v> 

" \ \ 



^^"^ v 


^C^ — 

"~^~^^ — * 

' — ■• 



3 5 4.0 4 5 5 3 5 4 4 5 5 3 5 
Junge Distribution Parometer - y 




Figure 5. — Temperature error AT versus the Junge distribution parameter v for 
a tropical atmosphere for three degrees of haziness and the 
particle size ranges indicated. The wavelength is 11.2 ym. 

Latitude Summer 




; 0.08 £ r £ lO.O^m 




40 45 50 35 40 4.5 50 35 4.0 4.5 5.0 
Junge Distribution Parameter - y 

Figure 6. — Same as figure 5, mid- latitude summer. 

2.0| — — i — — i — — i | 1 1 f 

Mid-Latitude Winter 



— 1.2 


5 1.0 


» 0.8 


• 0.6 


*- 0.4 





0.04 ^ r ^ 3.0// m 





_ 0.06 ^ r ^ 5.0/* r 


i i 



\ 0.08 -r 


\ V 

\ \ 
\ 1 

^ 10.0/um 








3.5 4.0 


50 3.5 4.0 4.5 5.0 3.5 


4.5 5.0 

Junge Distribution Parameter - if 

Figure 7. — Same as figure 5, mid-latitude winter. 


Subarctic Summer 





— Hazy 


— Clear 


^ r ^ 5.0/um 




s "**^CCj-- 










\ 0.08 ^ r 

^ lO.Oyum 

\ i 
\ i 
\ \ 
\ i 

-\ v 


\ x 




4.5 5.0 35 40 45 50 3.5 
Junge Distribution Parameter - 

Figure 8. — Same as figure 5, subarctic summer. 


45 5.0 


ic Wmtei 

- Clear 

- Hazy 




0.06 ^ 

r - 5.0// m 




5.0 3.5 40 4.5 5.0 3.5 4.0 
Junae Distribution Parometer - 1/ 

Figure 9. —Same as figure 5, subarctic winter. 


the larger size range is a serious limitation for cases requiring precise 
determinations of the surface temperature. 

Comparable statements can be made about the temperature errors for other 
atmospheric models, as shown in figures 6 to 9. In general, the magnitude 
of the errors decreases with increasing latitude, and the errors are smaller 
in winter than in summer for a given location and aerosol model. The reason 
for the trend toward smaller values of AT with decreasing surface temperature 
is not that the radiative fluxes are smaller, but that the lapse rates are 
more stable and the tropopause tends to occur at a lower level at the higher 
latitudes. Thus there is less difference between the temperatures of the 
aerosols and the surface at the middle and higher latitudes than in the 
tropics. In fact, the temperature inversion in the subarctic winter model 
causes T to be very small for that case (fig. 9) and the order of the curves 
is reversed so that AT decreases with increasing haziness of the atmosphere. 
The mechanism responsible for the reversal can be seen in figure 14. 

The data are segregated according to the haziness of the model In figures 
10 to 12. For the clear case (fig. 10), only for the largest size range and 
v < 4.0 does AT exceed 0.5°K, it is generally 0.2°K or less. As mentioned 
before, AT generally increases with increasing haziness and reaches nearly 
2.0°K for the worst conditions in a very hazy atmosphere (fig. 12). Only for 
the subarctic winter is increasing haziness accompanied by a decreasing 
magnitude of AT (compare figs. 10, II, and 12). 

The results so far have been for radiation emerging to space from the top 
of the atmosphere, but it is instructive to see what part of the atmosphere 
contributes most to AT. This feature is shown for the tropical atmosphere 
by the plot of AT vs. altitude in figure 13. The curves for all three size 
ranges exhibit the same general characteristics; only the magnitude of the 
temperature errors are different. Most of the error is produced in the highly 
concentrated aerosols near the surface. Above an altitude of about 4 km, the 
contributions to AT are all about the same for a given range of aerosol sizes. 
These characteristics are not unexpected, when the vertical profiles of 
aerosol concentrations are considered (fig. I). Only in the lowest levels of 
the atmosphere do the concentrations differ significantly among the models, 
so the primary differences in AT among them should occur in the lowest levels. 
At the higher altitudes the aerosol concentrations do not differ significantly 
for the models chosen, a fact which is responsible for the similar slopes of 
the curves above about 4 km. 

Figure 14 shows that the vertical profiles of AT for the mid latitude and 
subarctic summer models are qualitatively similar to those for the tropical 
model atmosphere, althoughthe magnitudes are somewhat different. In all of 
these cases most of the temperature error is introduced in the lowest layers 
of the atmosphere. In the subarctic winter model, on the other hand, a small 
negative value of AT is introduced in the lowest levels, with Its magnitude 
increasing as the haziness of the atmosphere increases. This reversal of 
sign is produced by the temperature inversion in the low levels of the sub- 
arctic winter temperature profile. Aerosol particles in the region from the 
surface to about 2 km altitude are at a temperature higher than that of the 
surface, and as a consequence add more radiation into the upward stream by 
emission than they subtract by absorption. Above the inversion, a greater 





1 4 







-I I- 

Mid-Latitude Summer 

-I I- 

Subarctic Summer 

-I I- 

Mid-Latitude Winter 

A V 

Subarctic Winter 

0.04 ^ r £ 3.0/im 

0.0 a?a ** tt w 






0.08 ^ 

r ^ 10.0/i 



^ i * iii ^^. 





3.5 4.0 4.5 5.0 3.5 4.0 4.5 50 3.5 4.0 
Junge Distribution Parameter - V 

45 5.0 

Figure 10. — Temperature error AT vs, the Junge distribution parameter v for 
a clear atmosphere for five atmospheric temperature profiles and 
the particle size ranges indicated. The wavelength is 11.2 ym. 



1 4 

1 2 


-\ 1 1 i 1 r 


H V 

Subarctic Summer 

H V 

Mid-Latitude Summer 

H V 

Mid-Latitude Winter 

H V 

Subarctic Winter 

0.04 £ r £ S.OyUi 







\ 0.08 ^ 

r £ 







- \% 






35 40 45 5.0 3.5 4.0 4.5 5.0 3.5 4.0 
Junge Distribution Parameter — V 

4.5 5.0 

Figure II. — Same as figure 10, hazy atmospnere. 

Very Hazy 



1 6 
1 A 





i 1 r 


-I I- 
Subarctic Summer 

h h 

Mid-Latitude Summer 

A r- 

Mid-Latitude Winter 

-I I- 
Subarctic Winter 

0.04 = r £ 3.0yu 

1 1 




v ■ 
\i . 
V : 

m °- 8 - r - 



- \ V 



- \\\ 



i i 


5 40 45 50 3.5 40 4.5 5.0 35 40 45 5.0 
Junge Distribution Parameter — y 

Figure 12. — Same as figure 10, very hazy atmosphere. 

Tropical, if = 4.0 

04 £ r £ 3.0p m 0.06 ^ r ^ 5.0// m 0.08 ^ r ^ 10.0^ m 

00 02 0.4 0.6 0.0 0.2 0.4 06 0.0 0.2 0.4 0.6 
Temperature Error (°K) 

Figure 1 3. —Temperature error AT vs. altitude for a tropical atmosphere for 
three degress of haziness, a single value of the distribution 
parameter v, and particle size ranges indicates. The wavelength 
is I I .2 ym. 

1/ =4.0, 0.08=r = 10.0/jim 


i W ' 
1 W 

i i | ii ; 

j ' I 1 /; I 




rl |i 

1 !! 

1 1 H 

1 Very j 1 ! ; 

Hazy ; I i ; 
1 1 I ! ; 


HI " 

1 1 1 : 


j i !/■ 

i : 1 ' : 


i if 


' I- 1 



! it 


1 i :/ 

i ill - 

1 it; Tropical 

' i/.i 

\W ] 



'I J L 

■ ll: 

■ \ ■ ; 

! ii 


l l| Summer 
1 if Mid LfltitM' 4 " 

I 1 !! 

i / ; i 


1 1 I : 


1 :f Summer 
| if Mid-Latitude 

1 ,'.: 

1 ;/ 

i / ': 

- I 

1 .': 
* / * ; 


I Winter 
t Subarctic / 


- / 

i Ii 



f V 




- I 


i i 

f i i 

>** i i 

-0.2 0.0 0.2 0.4 0.6-0.2 0.0 0.2 0.4 0.6-0 2 2 4 6 

Temperature Error (°K) 

Figure 14. — Temperature error AT vs. altitude for the size range 

.08 < r < 10.0 ym and the Junge distribution parameter v = 4.0 
for 3 degreees of haziness and 5 temperature profiles indicated. 

absorption than emission of energy by the particles causes the curves for 
the subarctic winter model to assume the same general character as that of 
the other models, but the negative values have to be compensated for by 
colder layers before the errors become positive. This results both in very 
small final errors of AT for the subarctic winter atmosphere, and in a de- 
crease of error with increasing haziness in contrast to the other cases. By 
a small increase of the temperature inversion in the model, the temperature 
error could be completely erased or reversed in sign for the subarctic winter 
atmosphere. Such an inversion is not at all unusual in subarctic or arctic 
regions, although it apparently does not occur as a mean condition. 


The figures were drafted by Robert Ryan, the manuscript was typed by 
Betty Loveless, and Paul Pellegrino computed and graphed most of the results. 


Dei rmendj fan, D., Electromagnetic Scattering on Spherical Polyd i spersions , 
American Elsevier Publishing Co., Inc., New York, N. Y., 1969, 290 pp. 

Elterman, L., "UV, visible and IR attenuation for altitudes to 50 km," 
Environmental research papers , No. 285, AFCRL-68-0 1 53, Optical Physics 
Laboratory, Air Force Cambridge Research Laboratories, Bedford, Mass., 
1968, 49 pp. 

Junge, C, "Die Ro I I e der Aerosole und der gasformigen Beimengungen der Luft 
im Spurenstoff hausha 1 1 der Troposphare, " Tel I us , 5, 1953, pp. 1-26. 

"The size distribution and aging of natural aerosols as determined from 
electrical and optical data on the atmosphere," Journal of Meteorology , 12, 
1955, pp. 13-25. 

Marlatt, W. E., Cole, H. I., Harlan, J. C, and Hjermstad, J. I., "Infrared 
radiation transfer through atmospheric haze layers," Final Report , NAS 
5-1 1631, Atmospheric I imiations to remote sensing, Department of Watershed 
Science, Colorado State University, Fort Collins, Colo., 1971, 62 pp. 

McClatchey, R. A., Fenn, R. W., Selby, J. E. A., Volz, F. E., and Garing, J. S., 
"Optical properties of the atmosphere (revised)". Environmental research 
papers , No. 354, AFCRL-7 1-0279, Optical Physics Laboratory, Air Force 
Cambridge Research Laboratories, Bedford, Mass., 1971, 85 pp. 

Rosen, James M., "Simu I taneous dust and ozone soundings over North and 
Central America", University of Minnesota-Atmospheric Physfcs-25, Report 
#N68-II4I4, 1967. 

Stowe, Larry L., Jr., "The effects of particulate matter on the radiance of 
terrestrial infrared radiation", Institute of Geophysics and Planetary 
Physics, University of California, Los Angeles, 1971, 109 pp. 

Zel 'manovich, I. S. and Shifrin, K. C, Tables of Light Scattering, Vol. IV , 
Leningrad, U.S.S.R., Hydrometeorolog ica I Office, 1971, 167 pp. 


(Continued from inside front cover) 

NESC 51 Application of Meteorological Satellite Data in Analysis and Forecasting. Ralph K. Anderson, 
Jerome P. Ashman, Fred Bittner, Golden R. Farr, Edward W. Ferguson, Vincent J. Oliver, and 
Arthur H. Smith, September 1969. Price $1.75 (AD-697-033) Supplement price $0.65 (AD-740- 

NESC 52 Data Reduction Processes for Spinning Flat-Plate Satellite-Borne Radiometers. Torrence H. 
MacDonald, July 1970. Price $0.50 (COM-71-00132) 

NESC 53 Archiving and Climatological Applications of Meteorological Satellite Data. John A. Leese, 
Arthur L. Booth, and Frederick A. Godshall, July 1970. Price $1.25 (COM-71-00076) 

NESC 54 Estimating Cloud Amount and Height From Satellite Infrared Radiation Data. P. Krishna Rao, 
July 1970. Price $0.25 (PB-194-685) 

NESC 56 Time-Longitude Sections of Tropical Cloudiness (December 1966-November 1967). J. M. Wallace, 
July 1970. Price $0.50 (COM-71-00131) 

NOAA Technical Reports 

NESS 55 The Use of Satellite-Observed Cloud Patterns in Northern Hemisphere 500-mb Numerical Analysis. 
Roland E. Nagle and Christopher M. Hayden, April 1971. Price $0.55 (COM-73-50262) 

NESS 57 Table of Scattering Function of Infrared Radiation for Water Clouds. Giichi Yamamoto, 
Masayuki Tanaka, and Shoji Asano, April 1971. Price $1.00 (COM-71-50312) 

NESS 58 The Airborne ITPR Brassboard Experiment. W. L. Smith, D. T. Hilleary, E. C. Baldwin, W. Jacob, 
H. Jacobowitz, G. Nelson, S. Soules, and D. Q. Wark, March 1972. Price $1.25 (COM-72-10557) 

NESS 59 Temperature Sounding From Satellites. S. Fritz, D. Q. Wark, H. E. Fleming, W. L. Smith, H. 
Jacobowitz, D. T. Hilleary, and J. C. Alishouse, July 1972. Price $0.55 (COM-72-50963) 

NESS 60 Satellite Measurements of Aerosol Backscattered Radiation From the Nimbus F Earth Radiation Ex- 
periment. H. Jacobowitz, W. L. Smith, and A. J. Drummond, August 1972. Price $0.25 (COM-72- 

NESS 61 The Measurement of Atmospheric Transmittance From Sun and Sky With an Infrared Vertical 
Sounder. W. L. Smith and H. B. Howell, September 1972. Price $0.30 (COM-73-50020) 

NESS 62 Proposed Calibration Target for the Visible Channel of a Satellite Radiometer. K. L. Coulson 
and H. Jacobowitz, October 1972. Price $0.35 (COM-73-10143) 

NESS 63 Verification of Operational SIRS B Temperature Retrievals. Harold J. Brodrick and Christopher 
M. Hayden, December 1972. Price $0.55 (COM-73-50279) 

NESS 64 Radiometric Techniques for Observing the Atmosphere From Aircraft. William L. Smith and Warren 
J. Jacob. January 1973. Price $0.35 (COM-73-50376) 

NESS 65 Satellite Infrared Soundings From NOAA Spacecraft. L. M. McMillin, D. Q. Wark, J.M. Siomkajlo, 
P. G. Abel, A. Werbowetzki, L. A. Lauritson, J. A. Pritchard, D. S. Crosby, H. M. Woolf, R. C. 
Luebbe, M. P. Weinreb, H. E. Fleming, F. E. Bittner, and C. M. Hayden, September 1973.